>>>>> On Sat, 28 Apr 2018 14:18:03 +0200, Thorsten Johannsen said: > > On 26.04.2018 18:00, Martin Simmons wrote: > >>>>>> On Thu, 26 Apr 2018 14:17:19 +0200, Thorsten Johannsen said: > >> > >> On 24.04.2018 18:13, Martin Simmons wrote: > >> > >> [...] > >> > >>> You could also kill the current bacula-fd process and start it under gdb, > >>> also > >>> adding the -d999 argument to verify that it still gets the segfault, e.g. > >>> > >>> sh$ gdb ...path.to.bacula-fd... > >>> (gdb) handle SIGSEGV > >>> (gdb) run -f -d999 > >>> > >> > >> Yes, this finally elicitated some useful information (well, just my > >> interpretation... some issue with strlen() ?? > >> > >> --------------------------------------------------------------------------------- > [...] > >> > >> Thread 3 "bacula-fd" received signal SIGSEGV, Segmentation fault. > >> [Switching to Thread 0x75d63450 (LWP 4550)] > >> strlen () at ../sysdeps/arm/armv6/strlen.S:26 > >> 26 ../sysdeps/arm/armv6/strlen.S: No such file or directory. > > > > Yes, this is a good... > > > >> (gdb) continue > > > > ...but this was not the correct command here (sorry I should have mentioned > > this) :-( > > > > Firstly, try to install the bacula dbg packages, e.g. > > > > sh$ apt-get install bacula-common-dbg bacula-fd-dbg > > > > Well, apparantly the debug packages are not available on Raspbian: > > -------------------------------------------------------------------------- > heRPI02:~ $ sudo apt-cache search bacula > > bacula - network backup service - metapackage > bacula-bscan - network backup service - bscan tool > bacula-client - network backup service - client metapackage > bacula-common - network backup service - common support files > bacula-common-mysql - network backup service - MySQL common files > bacula-common-pgsql - network backup service - PostgreSQL common files > bacula-common-sqlite3 - network backup service - SQLite v3 common files > bacula-console - network backup service - text console > bacula-console-qt - network backup service - Bacula Administration Tool > bacula-director - network backup service - Director daemon > bacula-director-mysql - network backup service - MySQL storage for Director > bacula-director-pgsql - network backup service - PostgreSQL storage for > Director > bacula-director-sqlite3 - network backup service - SQLite 3 storage for > Director > bacula-doc - Documentation for Bacula > bacula-fd - network backup service - file daemon > bacula-sd - network backup service - storage daemon > bacula-server - network backup service - server metapackage > > heRPI02:~ $ > -------------------------------------------------------------------------
Ah, OK. It you are running Debian stretch, then it looks like you have to add deb http://debug.mirrors.debian.org/debian-debug/ stretch-debug main to /etc/apt/sources.list and then run apt-get update to see the dbg packages. > > > > > > Then start it under gdb again with -f -d999 and use the following commands > > when it stops with the signal SIGSEGV: > > > > bt > > thread apply all bt > > info all > > x/i $pc > > x/64i $pc-0x30 > > disass htable::insert(char*, void*) > > > > > Nonetheless I tried your commands and got a lot of output this time. > I'm just not sure how relevant this information can be without the debug > information... > > > ------------------------------------------------------------------------- > heRPI02-fd: runscript.c:254-2019 runscript OK > heRPI02-fd: job.c:752-2019 Back from run_scripts ClientBeforeJob now: OK > heRPI02-fd: job.c:288-2019 <dird: accurate files=65860 > heRPI02-fd: job.c:311-2019 Executing Dir accurate files=65860 > command. > heRPI02-fd: htable.c:67-2019 malloc buf=752c5020 size=1000000 rem=435560 > heRPI02-fd: htable.c:130-2019 Leave hash_index hash=0x752c5068 > index=59779085 > heRPI02-fd: htable.c:283-2019 Insert: hash=0 index=59779085 > heRPI02-fd: htable.c:286-2019 Insert hp=752c5030 index=3015 > item=752c5030 offset=0 > heRPI02-fd: htable.c:293-2019 Insert hp->next=0 hp->hash=0x390280d > hp->key=<NULL> > heRPI02-fd: htable.c:299-2019 Leave insert index=3015 num_items=1 key=/mnt/ > heRPI02-fd: accurate.c:230-2019 add fname=</mnt/> lstat=LMC HtB EHt C A > A A BAA BAA I BaqFXr BaqEVw BaqFXr A A d delta_seq=0 chksum= > heRPI02-fd: htable.c:130-2019 Leave hash_index hash=0x6b8b0 > index=-1036830508 > heRPI02-fd: htable.c:283-2019 Insert: hash=3b2ac index=-1036830508 > heRPI02-fd: htable.c:286-2019 Insert hp=752c50b0 index=14281 > item=752c50b0 offset=0 > > Thread 3 "bacula-fd" received signal SIGSEGV, Segmentation fault. > [Switching to Thread 0x75d63450 (LWP 16098)] > strlen () at ../sysdeps/arm/armv6/strlen.S:26 > 26 ../sysdeps/arm/armv6/strlen.S: No such file or directory. > (gdb) > (gdb) bt > #0 strlen () at ../sysdeps/arm/armv6/strlen.S:26 > #1 0x76ef9a90 in bvsnprintf(char*, int, char const*, std::__va_list) () > from /usr/lib/bacula/libbac-7.4.4.so > #2 0x00000000 in ?? () > Backtrace stopped: previous frame identical to this frame (corrupt stack?) It looks like it is crashing while trying to print a debug message :-( I was beginning to suspect something like this, because several of the numbers in the debug output look bogus. This is caused by a bug in some of Bacula's debug messages. Can you try again with -d499 instead of -d999 which will avoid the htable.c messages? __Martin ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
